1. Field of the Invention
Apparatuses consistent with the present invention relate in general to a RF (Radio Frequency)-switch which allows an AC (alternating current) signal to pass therethrough by a bias voltage. More specifically, the present invention relates to a MEMS RF-switch using a semiconductor layer between a first electrode and a second electrode, thereby preventing charge buildup and sticking.
2. Description of the Related Art
Technical advances in MEMS (Micro Electro Mechanical System) have brought the development of a RF-switch based on the MEMS. In general, MEMS RF-switches have performance advantages over traditional semiconductor switches. For instance, the MEMS RF-switch provides extremely low insertion loss when the switch is on, and exhibits a high attenuation level when the switch is off. In contrast to semiconductor switches, the MEMS RF-switch features very low power consumption and a high frequency level (approximately 70 GHz).
The MEMS RF-switch has a MIM (Metal/Insulator/Metal) structure, that is, an insulator is sandwiched between two electrodes. Therefore, when a bias voltage is applied to the MEMS RF-switch, the switch acts as a capacitor, allowing an AC signal to pass therethrough.
FIG. 1 is a cross-sectional view of a related art MEMS RF-switch. As shown in FIG. 1, the MEMS RF-switch includes a substrate 11, a first electrode 12, an insulator 13, and a second electrode 15. Particularly, the MEMS RF-switch in FIG. 1 has a cap structure where the second electrode 15 packages the first electrode 12 and the insulator 13. Also, an air gap 14 exists between the second electrode 15 and the insulator 13.
When a bias voltage Vbias is applied in the direction shown in FIG. 1, the second electrode 15 is thermally expanded and shifts in the direction of the arrow, thereby making contact with the insulator 13. As such, the first electrode 12, the insulator 13 and the second electrode 15 act as a capacitor together, and the RF-switch is turned on, which in turn allows an RF signal to pass therethrough at a predetermined frequency band. However, if the bias voltage Vbias is not applied, the second electrode 15 shrinks and is separated from the insulator 13. As a result, the RF-switch is turned off and cannot allow the RF signal to pass therethrough.
When the bias voltage is applied, the second electrode 15 is charged positively resulting in a buildup of positive (+) charges, and the first electrode 12 is charged negatively resulting in a buildup of (−) charges. On the right hand side of FIG. 1 is a graph illustrating charges, or the quantities of electric charges, on the first electrode 12, the insulator 13 and the second electrode 15, respectively, of an ideal RF-switch. Referring to the graph in FIG. 1, the first electrode 12 which corresponds to the interval (0˜x1) is charged with −Qp, the second electrode 15 which corresponds to the interval (x3˜x4) is charged with +Qp. If the bias voltage is cut off in this state the charge turns back to 0. Meanwhile, the charge on the insulator 13 is maintained at 0, independent of the application of a bias voltage.
In practice, however, charge buildup often occurs to the insulator 13. Thus, the detected charge on the insulator 13 is not always 0.
FIGS. 2A and 2B are graphs for explaining charge buildup and sticking that occur to a non-ideal RF-switch. FIG. 2A illustrates a case when a bias voltage Vbias is applied. As shown, the first electrode 12 is charged with −Qp , the second electrode 15 is charged with +Q1 . At this time, +Q2 is built up on the insulator 13. Q1 and Q2 satisfy a relation of Q1+Q2=Qp. As such, although the bias voltage Vbias may be applied, a repulsive force is generated by the insulator 13 which is charged positively with +Q2 until the second electrode 12 is charged positively with greater than +Q2 . Therefore, the RF-switch is not turned on until a bias voltage with a certain magnitude is applied. As a consequence, switching time is increased.
Meanwhile, once the RF-switch is on, the insulator is charged with +Q2 and the first electrode 12 is charged with −Q2 even though the bias voltage may be cut off. As a result, sticking occurs because the second electrode 15 and the insulator 13 are not separated. Moreover, the RF-switch may not be turned off at all even when the bias voltage is completely cut off.